论文部分内容阅读
为量化驾驶人的驾驶适宜性,丰富对其的检测理论和方法,应用非集计理论中的多项分对数(MNL)模型构建驾驶适宜性度量模型。模型以驾驶人一定时间内事故发生次数作为选择肢,以个人固有属性、生理心理属性14项指标作为影响因素,并根据200份实际调查数据标定各影响因素参数。另外,选取60份数据验证该模型。结果显示:14项指标参数检验值均小于1.96,各参数统计学意义显著;模型判定系数为0.364 748,表明模型拟合程度较高;且该模型计算值与统计值最大绝对误差仅为3.3%,表明模型精度较高,可用于预测驾驶适宜性。
In order to quantify the driving suitability of drivers and enrich their detection theories and methods, a multi-item logarithmic (MNL) model of non-aggregate theory is applied to construct a driver suitability measurement model. The model takes the number of accident occurrences within a certain period of time as the choice limbs, and takes 14 indicators of personal inherent attributes and physiological and psychological attributes as the influencing factors, and calibrates the parameters of each influencing factor according to 200 actual survey data. In addition, select 60 data to validate the model. The results showed that the test values of 14 indexes were all less than 1.96, and the parameters of each parameter were significant. The coefficient of model determination was 0.364 748, which indicated that the fitting degree of the model was high. The maximum absolute error between the calculated value and the statistical value was only 3.3% , Indicating that the model has higher accuracy and can be used to predict driving suitability.